Abnormal metro passenger demand is predictable from alighting and boarding correlation
Bibliographic record
Abstract
Irregular sudden fluctuations in metro passenger demand during events or incidents can lead to critical supply or safety issues. Accurate and timely forecasting of such abnormal demand is crucial for effective crowd management and emergency response. However, this task remains challenging due to the absence of periodicity, high volatility, scarce samples, and the need for early warnings. This paper addresses abnormal metro passenger demand forecasting by leveraging the long-range Alighting-Boarding (AB) correlation driven by chained travel behavior. We propose a novel Alighting-Boarding Transformer (ABTransformer) model to explicitly capture the AB correlation with an interpretable bi-channel attention mechanism. Using real-world metro datasets from Guangzhou and Seoul, we demonstrate that leveraging the AB correlation significantly reduces the mean absolute error (MAE) over a six-hour forecast horizon by 5%–17% across three representative models. The ABTransformer performs best in forecasting abnormal metro boarding demand and remains competitive in normal demand forecasting. Notably, leveraging the AB correlation enables early warnings of abnormal demand with up to a 5-hour lead time (depending on the activity duration), offering an effective abnormal demand warning solution that does not rely on auxiliary event data. Additionally, we investigate uncertainty quantification in demand forecasting with different distribution assumptions. We observe multimodality in forecast distributions and find that simpler distributions, such as the zero-truncated Gaussian , tend to be more robust than complex mixture models in abnormal demand forecasting when observations are sparse. Our findings indicate that joint forecasting of alighting and boarding is always preferred over independent forecasting in metro passenger demand forecasting, particularly for abnormal demand scenarios.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".